5 Determination of the real refractive index

$Page 1: 5 Determination of the real refractive index$
Determination Of The Real Refractive Index

118

5 Determination of the real refractive index

5.1 Aim of study The results obtained from the previous study showed the importance of the refractive index, to

analyse results from laser diffractometry correctly. Unfortunately for most of the compounds

no refractive index can be obtained from the literature. Therefore in this chapter possibilities

for the determination of the refractive index were investigated. The real refractive index was

obtained for some selected nanosuspensions and NLC formulations in order to investigate to

which extent the refractive changes for the different systems. To investigate the usefulness of

these parameters for LD measurements, the optical parameters obtained were applied in LD

measurements and compared with the results from photon correlation spectroscopy and light

microscopy.

5.2 Methods to determine the real refractive index Literature screening gave view possibilities to determine the refractive index of solids and

suspensions, which are nicely described in (Malvern 2004; Rawle 2004; Rawle 2006). The

most appropriate methods where selected to be the observation of the Becke line and the

determination of the differential refractive index. All other methods e.g. the determination by

using polarisation microscopy or calculation via Gladestone Dale or Lorentz-Lorenz equation,

were not carried out. Either due to a lack of the required instruments (polarisation

microscope), other missing parameters (refractive indices of other polymorphs for Gladstone-

Dale approximation) or as the results were estimated to be too inaccurate (immersion method,

Lorentz-Lorenz equation).

5.2.1 Measurement of refractive index by analysis of dn/dc The dn/dc, also called differential refractive index is the variation of the real refractive index

due to a change in concentration of a solute. It is also known as the specific refractive index

increment with the given symbol ν. It is expressed as g/ml. However, mostly it is called dn/dc.

)(lim 0

00 c

nndcdn

cc

−==

→=

ν

The measurement of dn/dc gives the possibility to calculate the unknown index of refraction

of a compound. This is possible because the refractive index of a compound increases linearly

with an increase in its concentration, when it is dissolved or diluted in another medium.


119

The procedure and the calculation of the refractive index is described in the literature (Huglin

1972; Wu and Xia 1994; Russo 2005). The procedure used was kindly adapted from the

workgroup of R. Sigel (Sigel).

Different dilutions with various concentrations are prepared from the compound with the

unknown refractive index. Afterwards the real index of refraction is measured for each

solution prepared. Also a sample of pure dilution medium must be analysed. From the data

obtained the differential index of refraction can be analysed. For that the measured index of

the pure dissolution medium is subtracted from each index measured for the different

concentrations. The obtained set of data is plotted in a diagram; where the concentration is

plotted on the abscissa and the measured refractive index is plotted on the ordinate. From this

the linear regression function can be calculated. The slope m of the function corresponds to

the dn/dc in g/ml. In order to calculate the real refractive index of the unknown compound,

the obtained differential refractive index is multiplied by 100 and the refractive index of the

pure dissolution medium is added. The result corresponds to the real refractive index of the

compound. It is important to mention that the concentrations need to be measured in w/v% as

the differential refractive index is expressed in g/ml. However if the samples are weighted in

as volumetric concentration (v/v%), the value obtained needs to be corrected. For this the

obtained value for dn/dc has simply to be divided by the density of the compound with the

unknown refractive index. The advantage of the dn/dc determination method is that any

method for the measurement of the refractive index can be used (e.g. Abbe refractometer,

interferometer) and various parameters can be considered (different wavelengths and various

temperatures).

If particulate systems are analysed the concentration must be kept low, to avoid scattering

effects from particles. The particle size should be small for the same reason, as well as to

avoid sedimentation of particles during analysis. The determination coefficient of the

regression should be at least 0.99 or higher, as the result is very sensitive to errors, leading to

great variations in the calculated refractive index.

However, most of the systems analysed were not only suspended in water but also in

surfactant or glycerol/water mixtures or even in unknown suspension media (e.g. latex

suspensions). If the suspension medium is different to water, the pure dispersion medium

needs to be analysed as the reference. Otherwise the calculated index of refraction will be

incorrect. For instance a drug nanosuspension contains 1% of drug and is suspended in a 1%

surfactant solution. When the sample is diluted with water 1:1 also the concentration of the

surfactant is diluted 1:1. If no correction is considered the result would be a combination of


120

the refractive index of the particles and the surfactant. In Figure 5-1 an example is given for

the cyclosporine nanosuspension. Determination and dilution with water yields a refractive

index of 1.484, whereas the index is higher (1.497) when measured in the distinct stabiliser

solution, i.e. Poloxamer 188 0.5% in water.

y = 0.001510x + 1.332973R2 = 0.996444

y = 0.00163x + 1.33362R2 = 0.99659

1.33251.33301.33351.33401.33451.33501.33551.3360

0 0.2 0.4 0.6 0.8 1 1.2 1.4concentration in w/v%

refra

ctiv

e in

dex

n

cyclosporine nanosuspension analysed in watercyclosporine nanosuspension analysed in PLX 188

Figure 5-1: dn/dc for cyclosporine nanosuspension (NS) determined in water and PLX 188 solution Therefore those samples must be measured and diluted with the original suspension medium,

which can be prepared if the composition is known. If not, the procedure is more time

consuming, as particles need to be separated from the medium, to obtain pure dispersion

medium. In this study all dispersion media of the nanosuspensions and latex particles were

obtained by the separation of particles form the dispersion medium.

The dn/dc was determined using the Abbemat which determines the refractive index by

analyisis of the critical angle of total reflection. For the cylsosporine, ADA and

carbamazepine nanosuspensions the refractive index could also be analysed by using a scan

ref, which is based on interferrometry measurements.

Table 5-1 – Table 5-10 show the compositions of the systems analysed. The rational behind

selecting these particles was to have one group with identical stabiliser (lecithin) but differing

in the composition of the particle matrix (LCT; MCT: LCT mixture 50:50 and a wax). In a

second group the particles matrix is identical, but the stabiliser is different. In the next group

the incorporated drug was different. To change certain parameters of the particles should

allow assessing each specific contribution to the real refractive index.


121

Table 5-1: Fat emulsions and SLN - identical in stabiliser and dispersion medium 1 2 3 Lipofundin N 10% Lipofundin MCT 10% formulation C6 LC lipid (=SLN) compound w% compound w% compound w% soy bean oil (=LCT) 10.0 soy bean oil (=LCT) 5.0 cetylpalmitate (=wax) 10.0 egg lecithin 0.8 egg lecithin 0.8 egg lecithin 0.8 glycerol 2.5 glycerol 2.5 glycerol 2.5 medium chain

triglycerides (=MCT) 5.0

Table 5-2: Cetylpalmitate NLC - identical in lipid composition, different in stabiliser or incorporated drug 4 5 6 7 batch code: Stab. 1 with vitamin E

batch code: Stab. 1

batch code: Stab. 2

batch code: Stab. 3

compound w% compound w% compound w% compound w% cetylpalmitate 12.0 cetylpalmitate 15.0 cetylpalmitate 15.0 cetylpalmitate 15.0 Miglyol 812 4.0 Miglyol 812 5.0 Miglyol 812 5.0 Miglyol 812 5.0 Tegocare 450 1.8 Tegocare 450 1.8 Poloxamer 188 1.8 Tween 80 1.8 α-Tocopherol 4.0 Table 5-3: Softisan SLN - identical in lipid composition – different in incorporated drug 8 9 batch code: Softisan SLN placebo

batch code: Softisan SLN with liponic acid

compound w% compound w% Softisan 601 10.0 Softisan 601 9.5 Miranol Ultra C32 1.2 Miranol Ultra C32 1.2 liponic acid 0.5 Table 5-4: Stearyl alcohol NLC - identical in lipid composition – different in incorporated drug 10 11 12 batch code: stearly alcohol NLC /sunflower oil

batch code: stearly alcohol NLC placebo

batch code: stearly alcohol NLC with tretinoine

compound w% compound w% compound w% stearyl alcohol 10.5 stearyl alcohol 18.0 stearyl alcohol 17.73 sun flower oil 4.5 Miglyol 812 2.0 Miglyol 812 1.97 Tween 80 3.0 Tween 80 2.0 Tween 80 2.0 tretinoine 0.3 Table 5-5: Dynasan NLC 13 batch code: Dynasan compound w% Dynasan 116 20.0 Tyloxapol 5.0 Table 5-6: Nanosuspensions – different in drug – similar in stabiliser (Tween 80) 14 15 16 17 ADA (30.4.04 cycle 20 0°C RT ASLI)

budesonide (20.5.03 cycle 20 0°C RT ASLI)

carbamazepine (17.4.03 cycle 20 45°C RT ASLI)

itraconazole (28.1.03 cycle 20 10°C RT ASLI)

compound w% compound w% compound w% compound w% azodicarbonamide (ADA)

1.0 budesonide 1.0 carbamazepine 1.0 itraconazole 1.0

Tween 80 0.5 Tween 80 0.5 Tween 80 0.5 Tween 80 0.5


122

Table 5-7: Nanosuspensions – different in drug – similar in stabiliser (PLX 188) 18 19 20 21 bethamethason-valerat (BMV 01 25.8.04 Cycle 20 RT )

budesonide (BCH Z-40))

cyclosporine (C3-Spüli 5.3.04 KS)

buparvaqoune (xxx)

compound w% compound w% compound w% compound w% bethamethason-valerat

1.0 budesonide 1.0 cyclosporine 0.53 buparvaquone 1.0

PLX 188 1.0 PLX 188 1.0 PLX 188 0.25 PLX 188 0.5 Chitosan

chloride 0.5

Table 5-8: Latex dispersions 22 23 24 BMG 32 60/II SCA18 compound w% compound w% compound w% polysterene latex 7.2 polysterene latex 4.45 polysterene latex 4.7 PEG PEG PEG 200 Table 5-9: Nanosuspensions analysed with scan ref 25 26 20 cyclosporine (C3-Spüli 5.3.04 KS) carbamazepine (C4 Carb1 oG-

KS) azodicarbonamide (ADA) (C3- ADA 20n)

compound w% compound w% compound w% cyclosporine 0.53 carbamazepine 1.0 azodicarbonamide 1.0 PLX 188 0.25 PLX 188 0.5 PLX 188 0.5 Table 5-10. Various other compounds analysed Surfactants: SDS PLX 188 Tween 80 Oils: linseed oil miranol ultra Miglyol 812 almond oil peanut oil peppermint oil olive oil sun flower oil poppy seed oil soy bean oil corn oil saflor oil castor oil fish oil


123

5.2.1.1 Determined indices for selected lipid systems and nanosuspensions Table 5-11: Real refractive indices for various systems analysed with Abbemat

1.5001.497

1.5841.516

1.4931.489

1.5091.498

1.4421.510

1.5201.454

1.4971.548

1.5231.553

1.4511.549

1.5571.566

1.4971.492

1.4831.450

1.4561.479

1.4701.401

1.4491.460

1.4671.4701.4701.4731.4731.4731.4741.4741.4781.4791.480

1.400 1.450 1.500 1.550 1.600

Lipofundin 10% NLipofundin 10% MCT

SLN 10%Stab 2 (PLX 188)

Stab 3 (Tween 80)Stab 1 (Tegocare)Stab 1 with Vit.E

placeboliponic acid

placebotretinoine

sun flower oildynasan

azodicarbonamidecarbamazepine

itraconazolebuparvaquone

bethametasonvaleratbudesonide (Tw. 80)budeson. (PLX 188)

cyclosporinesca60/IISDS

glycerolPLX 188

Tween 80Miranol ultraMiglyol 812

peppermint oilolive oil

almond oilpeanut oil

soy bean oilcorn oil

sun flower oilpoppy seed oil

saflor oilcastor oil

fish oillinseed oil

cety

lpa

lmita

teso

fti-

san

stea

ryl

alco

hol

.

emul

-si

ons/

SLN

NLC

with

nano

susp

ensi

ons

late

xsu

rfact

ants

oils

real refractive index n

The results show, that the indices are different for each system analysed. It was not only

dependent on the main component, e.g. oil or drug, but also on the stabiliser used. Critically it

needs to mentioned that the concentrations used for the determination were not controlled by

an extra method e.g. HPLC (except for cyclosporine nanosuspension). It was assumed that the

concentrations remained constant over the whole homogenisation process when preparing the


124

dispersions. As this cannot be guaranteed the values obtained most likely are not the absolute

indices. However from the small standard deviations it can be seen that the method itself is

highly suitable for the determination of the real refractive index.

5.2.1.2 Limitations of the method For the system Latex BMG 22 irregular and negative refractive indices were obtained,

resulting in a refractive index smaller than water (Figure 5-2) As this is not logical the

obtained result was checked by using a different method to obtain information about the

refractive index (see 5.2.2)

y = -0.001021x + 1.331907R2 = 0.820689

1.3240

1.3260

1.3280

1.3300

1.3320

1.3340

0 1 2 3 4 5 6 7 8

refra

ctiv

e in

dex

n

Latex BMG Linear (Latex BMG)

Figure 5-2: Determined dn/dc for Latex BMG 22, would give a real refractive of 1.2298 which is smaller than the refractive index of water – the result is not correct

5.2.2 Analysis of real refractive index by observation of the Becke line The Becke line method is usually utilised to determine if the crystal or the particle of question

has a lower or higher index of refraction than the surrounding medium. The Becke line is a

band or a rim of light visible along a particle or crystal boundary in light.

A Becke line is the result of two facts. Both are related to refraction along the boundaries of

particles. First there is the fact that particles act as lenses as they tend to be thicker in the

centre and thinner towards the edges. Therefore, if the real index of refraction is higher as the

surrounding medium, the rays of the incident light converge towards the centre of the particle.

If the index of refraction is lower than the surrounding medium the rays converge towards the

edge of the particle. Internal reflection of the incident light occurs within the particle due to

the presence of vertical particle boundaries. These two effects concentrate the light into a thin

band within an object with a high index of refraction (1997)(1997)(1997)(Schmidt and

Heidermanns 1958; El-Hinnawi 1966; Richardson 1991; Stroiber and Morse 1994; 1997;

Derochette 2005).

In the previous chapter Latex BMG 22 was analysed by measuring dn/dc using an Abbemat

refractometer. The analysed results of the dn/dc measurement indicated a real index of


125

refraction smaller than water. The result is unexpected and there was a need to prove the

result obtained from this analysis. The investigation of the Becke line is time consuming, but

the advantage is a visible and therefore a doubt free result.

For the experiment the latex dispersion was diluted with water (1:50). The Becke line of the

dispersion was observed by light microscopy at higher and lower positions. In order to

compare the results to a standard, the Becke line was also investigated for air bubbles (lower

refractive index than water) and for dust particles (higher refractive index than water). The

images obtained are shown in Figure 5-3 for the standards and in Figure 5-4 for the Latex

dispersion.

Figure 5-3: Becke line of air bubbles and a small particle (attached to air bubble) in water (upper) and for dust particles (lower) at higher (left) and lower (right) positions under the microscope.

Figure 5-4: Becke line of latex particles, effects are the same as for dust particles


126

Air bubbles have a smaller index of refraction than water. Air bubbles are illuminated inside

at upper positions. If the position is lowered the objects becomes dark and a bright ring

around the object occurs. It is the opposite case if the object has a higher index of refraction

than the surrounding medium. This can be as seen for the dust particles and the small particles

next to the air bubbles.

When the latex dispersion was analysed, no effects were observed during microscopy, as the

particles were too small for the observation of the Becke line. However, the enlargement of

the images taken during microscopy analysis by computer led to a clearly visible result. The

latex particles are dark and surrounded by a bright line at higher positions. When the position

is lowered the particles become illuminated inside, which is similar to the dust particles. From

this it is concluded, that latex BMG has a higher index of refraction than water. It also gives

evidence that the results obtained from the previous analyses included an error. It is assumed

that the reason of the error observed is the particle size of the latex particles. The principle of

the Abbemat refractometer is the measurement of the critical angle, where total reflection

occurs. Therefore light intensity is measured under different angles. The angle where a

minimum of measured light intensity is observed corresponds to the critical angle. The

particle size of the latex dispersion BMG 22 is 1.7µm which is about 3 times higher than the

wave length used for the dn/dc measurement (632.8nm). With this ratio of particle size to

wavelength, the BMG 22 latex particles show Fraunhofer scattering characteristics.

Fraunhofer scattering is characterised by an intense forward scattering, but weak back

scattering, which means that the scattered intensity strongly depends on the angle of incident

light. Thus the scattering effects from the particles my disturb a correct detection of the

critical angle, as the minimum of light due to total reflection is overwhelmed by scattered

light.

5.2.3 Measurement of refractive index with manual Abbe refractometer Also the Abbe refractometer measures the index of refraction by the measurement of the

critical angle of total reflection. In contrast to the digital Abbemat refractometer, the

bright/dark field boarder is observed visually. Therefore, if disturbance due to light scattering

occurs it is expected to see an unsharp and fuzzy boarder line or even no boarder.

The samples listed in Table 5-12 were analysed using a manual ABBE refractometer (Carl

Zeiss, Jena). Table 5-12 also gives the result of the observations from the experiment. All

NLC/SLN systems gave a sharp line between the bright and the dark field. Latex particles

60II and SCA also gave a clear sharp line. When nanosuspensions were observed the line was


127

Table 5-12: Analysed samples and results of analysis NLC/SLN Stab 1 sharp line visible Vit E sharp line visible Stab 2 sharp line visible Stab 3 sharp line visible CL Lipid sharp line visible Dynasan sharp line visible latex sca18 sharp line visible 60II sharp line visible BMG 22 no line nanocrystals budesonide sharp line visible, some fuzziness itraconazole sharp line visible, some fuzziness carbamazepine sharp line visible, some fuzziness ADA sharp line visible, some fuzziness cyclosporine sharp line visible, some fuzziness

existent, but some fuzziness was visible. No line was visible when latex particles BMG 22

were analysed. In order to correlate the results from Table 5-12 with the size all samples were

analysed by PCS and laser diffractometry. UV/Vis spectroscopy was also performed. UV/Vis

spectroscopy is mainly used to measure the absorption of compounds dissolved in a liquid.

From the determined absorption the concentration can be calculated according to Lambert

Beer`s law. If particular systems are analysed possible scattering phenomena from the

particles need to be considered. According to the scattering phenomena of particles (see 4.1),

particles larger than the wavelength will have stronger forward scattering than particles

smaller than the wavelength. Differences in the detected intensities are also expected if a

particle dispersion (constant size and concentration) is analysed with different wavelengths. If

one decreases the ratio between the particle size and the wavelength the forward scattering of

the particle increases. This leads to a higher intensity on the detector (determined as

transmission) and a smaller absorption therefore. The rational behind the UV/Vis

measurements of this study was to gain information about possible differences in the wave

length dependent scattering of the particles. It was hoped to see wavelength dependent

increases or decreases for the different particles and particle sizes. The dilution of the particle

dispersions in the same manner seemed to be difficult, because all dispersions were different

in particle size and particle concentration. On order to obtain a standard, the concentration of

the particle dispersions was adjusted using the laser diffractometer LS 230. This is possible as

the instrument measures the obscuration of the sample in the PIDS cell prior the

measurement. The obscuration in principle is the measurement of the transmission, which

means all samples were diluted in order to obtain the same transmission at 450nm (= PIDS

obscuration). The samples were obtained by preparing the LS 230 for a measurement with

included PIDS. After alignment, background measurement and the de-bubbling procedure


128

each sample was added to the sample cell until a PIDS obscuration of 45% was reached. The

cell was flushed and the diluted sample collected and analysed by UV/Vis spectroscopy

immediately. Figure 5-5 and Figure 5-6 show the data obtained from the measurements.

0.70.750.8

0.850.9

0.951

1.051.1

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

Stab 1 Vit E Stab 2 Stab3 CL Lipid Dynasan sca18 60II BMG 22 Budes. Itrac. Carba. ADA Cyclos.

NLC/SLN Latex Nanocrystals

Tran

smis

sion

/Inte

nsity

1003005007009001100130015001700

z-A

vera

ge in

nm

UV Transmission at 900,750,600 and 450nm PCS Figure 5-5: UV/Vis Transmission for 450, 600, 750 and 900nm and PCS values for the systems

00.020.040.060.080.1

0.120.140.160.180.2

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

450n

m60

0nm

750n

m90

0nm

Stab 1 Vit E Stab 2 Stab3 CL Lipid Dynasan sca18 60II BMG 22 Budes. Itrac. Carba. ADA Cyclos.

NLC/SLN Latex Nanocrystals

Abs

orpt

ion

1003005007009001100130015001700

z-A

vera

ge in

nm

UV Absorption at 900,750,600 and 450nm PCS Figure 5-6: UV/Vis Absorption for 450, 600, 750 and 900nm and PCS values for the systems In all systems the transmission increased, when the wave length was increased. Only latex

(BMG 22) showed a trend different to the systems. Here the transmission decreased when the

wave length was increased, as it was expected to see.

In conclusion, two effects lead to the absorption and transmission values measured. If

particles are present in the sample analysed, they interfere with the light beam of the

instrument in a size dependent manner. If particles are small and the wavelength is large, the

interference is small. If particles are about the same size than the wavelength or larger, the

interference is larger too. If the size is kept constant and the wavelength is increased, the

interference of light with the particles decreases with an increase of wavelength. This effect

was seen for all samples, expect for Latex dispersion BMG 22. Here the second and expected

effect could be observed. If light co-interacts with a particle scattering effects occur (e.g.

diffraction, refraction, reflection). Also here the effects are size dependent. The larger the

particle size in comparison to the incident wavelength, the more intense is the forward

scattered light intensity. This means with an increase in particle size the detected light


129

intensity from forward scattered light increases in a wavelength dependent manner. At a

certain particle size the intensity from scattering overwhelms the effect of absorption and

interference measured for smaller particles. In case of the latex particles the absorption value

measured for 450nm was even negative, clearly indicating a higher intensity on the detector

than in the medium without particles which derived from scattering of the particles.

From this it is concluded, that the refractive index of particles much larger than the

wavelength of the refractometer used can not be analysed due to scattering effects, which

disturb the detection of the total reflection, if disturbance occurs can be investigated by using

a manual Abbe refractometer. If the line between dark and broad field is sharp, the sample can

be analysed. UV/Vis spectroscopy also gives information about scattering effects. If a sample

can be analysed, the absorption values of the sample measured must decrease with an increase

in wavelength. If the trend is an increase of absorption with an increase in wavelength

scattering effects from particles are too intense and overwhelm other effects, e.g. refraction,

making an analysis impossible. Therefore particles to be analysed should be as small as

possible to keep the effect of scattering low.

The investigation of the Becke line is possible. Only particles larger 1µm can be analysed, as

the enlargement of the light microscope is limited. This method only gives a rough result of

the refractive index. The advantage is the visible and doubt free result therefore. Of course to

approach the real refractive index dispersion media for the particles need to be used having

different refractive indices. By determining whether the refractive index of the representative

liquid is higher or lower, one can stepwise approach the particle index.

5.2.4 Real refractive index obtained from interferometry In Figure 5-7 the indices which were obtained from the measurement with the Abbemat and

the scan ref are compared.

1.497

1.496

1.49521.49541.49561.49581.496

1.49621.49641.49661.49681.497

1.4972

cyclosporine

refra

ctiv

e in

dex

n

abbematscanref

Figure 5-7: Real refractive indices obtained for cyclosporine nanosuspension with scan ref (green) and Abbemat (blue)


130

The results are almost the same and differ only in the third digit after the dot. However the

index of refraction is dependent on the temperature and the wavelength and decreases if both

are increased. The Abbemat analyses the refractive index at the sodium wavelength

(589.3nm) whereas the scan ref is operated at a wavelength of 632.8nm. Therefore the

difference can be explained by the different wavelengths applied. In addition, for the particle

size analysis it needs to be considered that the LS 230 operates with a laser having a

wavelength of 750nm. Hence the index of refraction measured at a lower wavelength is higher

if the sample is analysed in the LS 230. Moreover it was found that the temperature, at which

the measurement takes place in the LS 230, is not - as probably assumed - 20°C but higher up

to 31°C. Therefore the refractive index further decreases.

5.2.5 Changes of the real index of refraction due to hydration Within the study also liquid polyethylene glycols (PEGs) have been analysed, by determining

dn/dc, primarily to prove the accuracy of the method. However the refractive indices which

were obtained from dn/dc calculations were not in agreement with the indices obtained from

the pure liquids and from the data found in the literature. All indices analysed were smaller as

expected (Figure 5-8).

1.460

1.456 1.456

1.448

1.4651.463 1.463

1.453

1.468

1.464 1.464

1.453

1.435

1.44

1.445

1.45

1.455

1.46

1.465

1.47

100%

tota

l*

high

**

low

***

100%

tota

l*

high

**

low

***

100%

tota

l*

high

**

low

***

PEG 200 PEG 400 PEG 600

refra

ctiv

e in

dex

n

0.998

0.9985

0.999

0.9995

1

R2

* analysed with complete set of data;**measured using only data from concentrations higher 5%*** anlysed using only data from concentrations lower 5%

Figure 5-8: Refractive indices for various polyethylene glycols obtained from pure liquid (= 100%) and determined from dn/dc measurements


131

The observation can be explained by the fact that PEGs show strong hydration phenomena

when dissolved in water, which can be observed by measuring the changes in temperature.

For the experiment 50.00g PEG were weighted in into a 100ml volumetric flask and filled up

with water. The temperature was determined at its highest point using a digital thermometer

(IKA-TRON DTM 10, Janke and Kunkel, Staufen, Germany). The following dilutions (25%,

12.5%, 6.25%, and 3.125%) were obtained by placing 50.0ml of the previous dilution into a

100ml volumetric flask and filling up the flask with water. The temperatures were measured

as described above. The data obtained are shown in Figure 5-9. The temperature increases if

water is added to the PEGs. With ongoing dilution the changes in temperatures decrease. At a

concentration of 3.125 %, no differences can be measured anymore when more water is

added. Therefore this concentration was thought to be related to the complete hydration of the

PEGs.

0

2

4

6

8

10

12

50 25 12.5 6.25 3.125

concentration of PEG in w/v% .

∆T in

°C

PEG 200PEG 300PEG 600

Figure 5-9: Increase in temperature due to hydration of the PEGs From this the data of dn/dc measurements were analysed by only calculating the refractive

index from the data of higher concentrations (>5%) and lower concentrations (< 5%), from

the latter it was assumed that PEGs were fully hydrated (Figure 5-8, orange and light green

columns). By comparing the determination coefficients (R2), it was found that it was higher

when fully hydrated PEGs (low concentrations) were analysed, but it was low when only

partly PEGs (higher concentrations) were analysed. Which means only low concentrations,

were no changes in temperature were observed, yielded accurate and therefore correct results

in principle, whereas concentrations were the temperature was still increasing upon the

addition of water, give non-linear results. The refractive indices obtained from low


132

concentrations are much smaller than those from the pure liquids. In conclusion: hydration of

PEGs decreases the real refractive index. The decrease due to hydration was about 0.015 in

total. Even if the difference seems little, it should be considered.

5.2.6 Time dependent changes in refractive index The dn/dc of a cyclosporine nanosuspension was analysed at different times after the

production. The results which are shown in Figure 5-10 indicate that the refractive index

increases over the time of storage. Any loss of dilution medium due to evaporation could be

excluded by the determination of the total concentration of cyclosporine via HPLC. Therefore

it is assumed that the index changes due to a change in the crystalline state of the nanocrystals

(e.g. transformation from amorphous to crystalline).

1.476

1.484

1.4931.497

1.515

1.45

1.46

1.47

1.48

1.49

1.5

1.51

1.52

0 20 30 51 95

age of nanosuspension in days

refra

ctiv

e in

dex

Figure 5-10: Changes in refractive index of cyclosporine nanosuspension over time

5.2.7 Conclusion In this study a method to determine the unknown refractive index for nanosuspensions and

NLCs was established. The method determines the refractive index by the measurement of the

specific refractive increment (dn/dc). It was shown, that the index of refraction is dependent

on the main compound but also on the stabiliser or co-compounds. The method is not suitable

for larger particles, as scattering caused by the interference disturbs the correct function of the

instrument. The observation of the Becke line can be used to check doubtful results, but is not

reliable for the detailed determination of the refractive index. Also a manual Abbe

refractometer is not the appropriate instrument, as the resolution is only 0.002, which is too

inaccurate for the determination of dn/dc, where mostly low concentration with little

differences in RI are measured. However the Abbe refractometer was found to be useful if the


133

all over possibility for the determination of dn/dc needs to be checked. Only particles giving

a sharp border can be analysed. Any fuzziness or no border indicates that particles are too

large for analysis.

The refractive index is also influenced by the hydration effect, as studied for PEGs in this

study. Hydrations were found to decrease the refractive index. More interestingly it was found

that the refractive index of a cyclosporine nanosuspension increased over the time of storage.

This observation should be investigated in more detail, as it may open a new perspective to

observe changes in the crystalline state, which is up to now only possible by less convenient

x-ray diffraction. Refractive index measurements might even be able to detect structural

changes in the surface layer of nanoparticles which are not accessible by x-ray due to the

resolution limit (5% of a fraction must be present in the sample to be detected).

In summary refractive indices can be analysed. However, as the refractive index strongly

depends on the temperature and the wavelength, it would be more accurate to obtain the

required refractive index at a wavelength similar to the one used for LD measurements. In the

case of the LS 230 this is not possible, as no instrument operated at this wavelength is

available on the market. Also further investigations should be carried out at temperatures

similar to the temperature of the actual measurement. This was not taken into account here, as

the measurements were conducted before the unexpected increased temperatures within LD

measurements were realised.

5.3 Comparison of LD results obtained from measurements with and without correct optical parameters

In order to compare the meaning of correct optical parameters in practice, the systems from

which the real refractive index was determined were analysed by laser diffractometry. For

comparison the raw data were calculated using Fraunhofer approximation and Mie modus,

where the standard value of 1.456 (Müller 1996) was compared to the definite determined

values. Values for the imaginary indices were analysed as suggested by Beckman Coulter by

simply determining UV/VIS absorption of the original sample in the dispersion medium. In

addition all systems were characterised using PCS and light microscopy. In the following

tables all data for each system are put together.


134

In Table 5-13 the data obtained for the Lipofundin MCT emulsion are shown. Microscopic

imaging detected some droplets larger than 2µm, which were not found within LD

measurements, neither by Fraunhofer approximation nor by both Mie calculations. The

simulation gives particle sizes larger than 1µm only for high IRI values. The graph shown in

the third row of the table shows the particle size distribution calculated. In this Fraunhofer

detects a very small particle population of approximately 50nm, indicating the poor sensitivity

of the calculation mode. Nevertheless small fractions of about 80nm were detected for both of

the Mie calculations. If this peak is correct is hardly to decide, but might be alright as it is

known, that unbound lecithin can form liposomes, being in this size. All over in this analysis

the differences between the Mie calculations are neglectable small, whereas Fraunhofer

approximation failed.

In Table 5-14 the data from the Lipofundin N 10% emulsion are shown. From microscopy no

larger droplets were obtained, matching with the results from laser diffractometry therefore.

Again Fraunhofer approximation failed, as the small peak already seen in the MCT emulsion

also oocurs here. In contrast to the MCT formulation, Mie with the standard refractive index

detected a smaller peak, thought to correspond to liposomes, whereas the refractive index

determined did not detect the second particle population. Also here it is hard to decide which

mode matches the real result best. The result only shows that a change of the refractive index

changes the particle size and the size distribution. Depending on which mode used the

interpretation of the results will vary, e.g. if the determined index is used the MCT emulsion

would be thought to be more instable, as liposomes (second peak) were detected in

comparison to the N formulation were no liposomes could be observed. In contrast; Mie mode

with the standard value gives a second peak for both of the emulsion - here a differentiation is

actually not possible. However, Fraunhofer gave incorrect results for both of the systems.

The data obtained for Stab 1 (NLC containing cetylpalimate, Miglyol 812 and Tegocare) are

shown in Table 5-15. The analysis led to the same result. Fraunhofer gave a totally different

size and size distribution as the Mie calculations, which were almost the same. However the

simulation showed that the use of higher imaginary values (i.e. 1.0!) increases the particle size

detected. In the distribution graph (third row) it is analysed as a third particle population at

about 2µm, which was also seen from microscopy. However such high values are not correct.

Therefore the matching result of that optical parameter with microscopy probably just

occurred by chance. The lack of the detection of larger particles is thought to be related to the

fact as already demonstrated. Larger particles are often not detected if PIDS is included into

the measurement.


135

In contrast was Stab 2 (NLC as Stab 1 but stabilised with Poloxamer 188) (Table 5-16). As

seen from microscopy the system consisted of large agglomerates and fat particles, whereas

only a few nanoparticles were left. Large particles were detected in every calculation mode

applied. However, Fraunhofer in this case failed to detect smaller particles. When Mie modus

with optical parameters from (Müller 1996) was used the second peak at approx. 2µm is not

found. Larger particles than 500µm are not detected in this modus, but in the Mie mode with

correct optical parameters. All over, all analyses give the result of the existence of larger

particles, but the Mie analysis with the determined index is best, as it resolves small particles,

as well as it detects larger particles.

The data of Stab 3 (NLC as Stab1 and 2 but stabilised with Tween 80) (Table 5-17) also

showed large particles in microscopy as well as in all LD calculations. The differences of the

Mie calculations are little.

Fraunhofer analysis for the SLN stabilised with lecithin (Table 5-18) detected particles up to

3.206µm not found in images from microscopy. Also particles smaller than 60nm were

detected incorrectly. Mie analysis after (Müller 1996) finds particles up to 869nm, whereas

the analysis with the correct index of refraction detect particles up to 1.047µm, which were

also seen from microscopy. The tendency is that Mie analysis after (Müller 1996) leads

preferentially to too small particle sizes.

For the Dynasan NLCs, which are shown in Table 5-19, Fraunhofer found particles up to

3.205µm. In both Mie analysis no larger particles than 452nm were detected, but were seen in

images from microscopy. The results indicate that a correct index of refraction is not the only

important parameter for a correct analysis.

The stearyl aclcohol NLCs without tretinoine (Table 5-20) contained large particles, as

detected from all LD measurements. In this case the analysis by microscopy did not show any

larger particles. From this it is shown again, that also microscopy can lead to errors of

analysis.

The analysis of NLCs containing tretinoine (Table 5-21) detected no particles larger 3.5µm

within LD analysis. Also microscopy showed only a single crystal of 30µm.


136

Table 5-13: Data of Lipofundin MCT 10% Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

5µm

magnification: 1000x

00.10.20.30.40.50.60.70.8

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

Lipofundin MCT 10%

parti

cle

size

in µ

m

Fraunhofer Mie after (Müller 1996)

Mie with 1.497 and 0.01 PCS z-Average

0

0.2

0.4

0.6

0.8

1

1.2

00.

001

0.01

0.03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1

1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

Lipofundin MCT 10%

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


137

Table 5-14: Data of Lipofundin N 10% Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

5µm


00.2

0.4

0.6

0.8

1

1.2

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

Lipofundin N 10%

parti

cle

size

in µ

m



00.20.40.60.8

11.21.41.61.8

00.

001

0.1

0.03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1

1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

Lipofundin N 10%

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


138

Table 5-15: Data of Stab 1 (lipid phase: cetylpalmitate and Miglyol 812, stabiliser: Tegocare 450) Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

5µm


0

0.2

0.4

0.6

0.8

1

1.2

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

Stab 1

parti

cle

size

in µ

m



0

0.51

1.52

2.53

3.5

00.

010.

030.

05 0.1

0.3

0.5 1 3 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 3 0

0.01

0.03

0.05 0.1

0.3

0.5 1 3 0

0.01

0.03

0.05 0.1

0.3

0.5 1 3 0

0.01

0.03

0.05 0.1

0.3

0.5 1 3 0

0.01

0.03

0.05 0.1

0.3

0.5 1 3 0

0.01

0.03

0.05 0.1

0.3

0.5 1 3 0

0.01

0.03

0.05 0.1

0.3

0.5 1

1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

Stab 1

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


139

Table 5-16: Data of Stab 2 (lipid phase: cetylpalmitate and Miglyol 812, stabiliser: Poloxamer 188) Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

5µm


050

100150200250300350400450

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

Stab 2

parti

cle

size

in µ

m



0

50

100

150

200

250

300

00.

010.

030.

05 0.1

0.3

0.5 1 3 0

0.01

0.03

0.05 0.1

0.3

0.5 1 3 0

0.01

0.03

0.05 0.1

0.3

0.5 1 3 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1

1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

Stab 2

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


140

Table 5-17: Data of Stab 3(lipid phase: cetylpalmitate and Miglyol 812, stabiliser: Tween 80) Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

5µm


0

50

100

150

200

250

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

Stab 3

parti

cle

size

in µ

m

Fraunhofer Mie after (Müller 1996)Mie with 1.493 and 0.04 PCS z-Average

0

50

100

150

200

250

00.

010.

030.

05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1

1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

Stab 3

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


141

Table 5-18: Data of cetylpalimitate SLN stabilised with lecithin Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

5µm


00.5

11.5

22.5

33.5

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

cetyl palmitate SLN stabilized with lecithin

parti

cle

size

in µ

m



00.5

11.5

22.5

33.5

00.

001

0.01

0.03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1

1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

lecithine SLN

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


142

Table 5-19: Data of Dynasan NLC Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

5µm


00.5

11.5

22.5

33.5

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

Dynasan NLC

parti

cle

size

in µ

m

Fraunhofer Mie after (Müller 1996)Mie with 1.497 and 0.02 PCS z-Average

0

0.51

1.52

2.53

3.5

00.

001

0.1

0.03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1 0

0.00

10.

10.

03 0.1 1

1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

Dynasan NLC

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


143

Table 5-20: Data of stearyl alcohol NLC without tretinoine Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

50µm

magnification: 160x

050

100150200250300350400

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

Stearyl alcohol NLC without tretinoine

parti

cle

size

in µ

m



050

100150200250300350400

00.

010.

030.

05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1

1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

Stearyl alcohol NLC

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


144

Table 5-21: Data of stearyl alcohol NLC containing the drug tretinoine Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

50µm

magnification: 160x

00.5

11.5

22.5

33.5

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

Stearyl alcohol NLC with tretinoine

parti

cle

size

in µ

m



00.5

11.5

22.5

33.5

00.

001

0.01

0.03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1

1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.6 1.7 1.75 1.8

Tretinoine NLC

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


145

Table 5-22: Data of Azodicarbonamide nanosuspension Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

50µm

magnification: 160x

01020304050607080

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

ADA

parti

cle

size

in µ

m



0102030405060708090

00.

001

0.01

0.03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1 1 0

0.00

10.

010.

03 0.1

1.3 1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75

azodicarbonamide nanosuspension

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


146

Table 5-23: Data of cyclosporine nanosuspension Comparison of PCS diameter (z-average ) and LD diameters calculated with Fraunhofer (blue) and Mie using the standard values after (Müller 1996) (orange) and the measured RI and IRI (green) (upper right), the corresponding LD distribution curves (middle), the corresponding microscopic picture (upper left), the full simulation of the LD raw data (using optical parameters ranging from 1.3-1.8 for RI and from 0-1 for IRI) (lower), LD measurement was performed with included PIDS

5µm


0123456789

10

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100

cyclosporine

parti

cle

size

in µ

m



0123

4567

00.

010.

030.

05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1 0

0.01

0.03

0.05 0.1

0.3

0.5 1

1.35 1.4 1.45 1.5 1.55 1.6 1.65 1.7 1.75 1.8

cyclosporine nanosuspension

size

in µ

m

LD 10 LD 50 LD 90 LD 95 LD 99 LD 100


147

However this NLC suspension was known to be highly polydisperse. Particle size

characterisation in this way leads to fatal errors, even though if the optical parameters are

correct. The reason was found to be the inclusion of PIDS into the measurement, which over

estimates small particles. Therefore this system as well as dynasan SLN and the cyclosporine

nanosuspension were analysed again without PIDS. For all systems now larger particles were

clearly detected (Figure 5-11).

Figure 5-11: LD data without PIDS from Dynasan SLN, Tretinoine SLN and cyclosporine nanosuspension

The analysis of the nanosuspension containing azodicarbonamide by light microscopy showed

a highly agglomerated system (Table 5-22). This was not detected when Fraunhofer

approximation was applied. If Mie theory was applied, larger particles could be analysed.

However, differences can be seen in the distribution graph. Mie theory with the standard

value gives maximal particle sizes of only 40µm, whereas the maximal value was 80µm if the

determined correct refractive index was used. The analysis of the cyclosporine

nanosuspension (Table 5-23) shows a trimodale distribution for the system if Fraunhofer

approximation is used. Mie calculations were very similar to each other and only found a

monomodal distribution. In the LD chapter the performance of the LS 230 was investigated

by using various mixtures of latex dispersions. None of the results was analysed correctly,

when Fraunhofer approximation was applied. Therefore also the result obtained from

Fraunhofer analysis for the cyclosporine nanosuspension is expected to be an artificial

distribution.


148

5.3.1 Conclusion The comparison of all results obtained from these measurements gave very distinct results. In

some of the systems analysed, large particles were not found, either by microscopy or/and by

laser diffractometry.

Fraunhofer approximation often failed to characterise the system correctly. The comparison of

both Mie calculations (standard after (Müller 1996) vrs. correctly determined index) also gave

different results. In most of the cases the difference between the results was only very small,

whereas in some of the cases very different results were obtained. From the simulations,

where various optical parameters were used it was found that every system is influenced

differently by these parameters. Some systems gave stronger variations than others. This was

also found for the size distribution curves. Some size distributions changed from bimodal to

monomodal or trimodal. However taking into account the findings from the nailing test, those

distributions are not necessarily correct or true. In conclusion the LS 230 is not the

appropriate instrument for the detection of polydisperse but narrow systems in the submicron

range. Therefore the interpretation of calculated polydisperse results should be done with

caution. All over it is more important to measure the systems also without PIDS, as only here

larger particles are detected reliably.

Documents

5 Determination of the real refractive index